Загрузил mitrosem2001

matecconf sg2015 01039

реклама
See discussions, stats, and author profiles for this publication at: https://www.researchgate.net/publication/287974469
Using Static Polynomial Load Models in “Rastrwin” Software
Package for Power System Studies
Article in MATEC Web of Conferences · January 2015
DOI: 10.1051/matecconf/20153701039
CITATIONS
READS
2
53
3 authors:
Aleksey V. Pankratov
Mikhail A. Kondrashov
Tomsk Polytechnic University
Tomsk Polytechnic University
8 PUBLICATIONS 17 CITATIONS
6 PUBLICATIONS 4 CITATIONS
SEE PROFILE
SEE PROFILE
Svetlana Paul
Tomsk State University
17 PUBLICATIONS 37 CITATIONS
SEE PROFILE
Some of the authors of this publication are also working on these related projects:
Identification of the static load characteristics of the remote metering data View project
All content following this page was uploaded by Aleksey V. Pankratov on 08 February 2017.
The user has requested enhancement of the downloaded file.
MATEC Web of Conferences 3 7, 0 1 0 39 (2015)
DOI: 10.1051/matecconf/2015 370 103 9
C Owned
by the authors, published by EDP Sciences, 2015
USING
STATIC
POLYNOMIAL
LOAD
MODELS
IN
“RASTRWIN” SOFTWARE PACKAGE FOR POWER SYSTEM
STUDIES
1
1a
2
Aleksey Pankratov , Mikhail Kondrashov , Svetlana Paul
1
National Research Tomsk Polytechnic University, 634050 Tomsk, Russia
National Research Tomsk State University, 634050 Tomsk, Russia
2
Abstract. Polynomial load model practical application has some specificity. Per unit static load
model dependence on node rated voltage value is demonstrated in this paper. The assigned
rated voltage values of load nodes with polynomial characteristics are to be invariant to avoid
calculation complementary error appearance. If rated voltage correction is necessary condition,
for example to obtain calculation convergence, the active and reactive power drawn under rated
voltage and polynomial coefficients must be corrected too. Correcting calculation expressions
are presented in the paper.
1 Introduction
Load modeling is necessary for power system management. Different load model modifications are established
[1-3], but problems of its practical application are described not so particularly.
The basic simulation tool, using by the System Operator of the Russian United Power System, is “RastrWin”
software package [4]. Electric loads are presented by per unit polynomial static load models in this package. Base
voltage, active and reactive power are used to convert values to per unit system.
Practical calculations of power system mode parameters often demand nodes base voltage changing. In result
static load models have difference from original characteristics, and calculation complementary error appears.
There is no description of this problem and its solution in literature.
2 Experimental setup and study technique
Simple radial network, supplying cable factory “Sibkabel” of Tomsk city, was modeled in the
“RastrWin” software package. Static load models in “RastrWin” are described by expressions
a
Corresponding author: mihail-arhangel0@mail.ru
4 Article available at http://www.matec-conferences.org or http://dx.doi.org/10.1051/matecconf/20153701039
MATEC Web of Conferences
2
U U
P(U ) PBAS a0 a1 a2 ,
U BAS
U BAS (1)
2
U U
Q(U ) QBAS b0 b1 b2 ,
U BAS
U BAS where PBAS , QBAS , U BAS – base active and reactive power and base voltage; a0 , a1 , a2 , b0 , b1 , b2 –
per unit polynomial coefficients. Load node rated voltage is used as base voltage. But rated voltage
term differs from classical interpretation in “RastrWin” software package. Thus using node rated
voltage values can vary from the standardized quantities.
Original static load model parameters were obtained in active experiment with base voltage
U BAS1 6.2 kV and are presented in Table 1.
Table 1. Polynomial load model coefficients, base active and reactive power with U BAS1 6.2 kV
Active power
Reactive power
a0(1)
a1(1)
a2(1)
PBAS(1) , MW
b0(1)
b1(1)
b2(1)
QBAS(1) , Mvar
1.481
-2.277
1.796
1.4
52.048
-108.651
57.599
0.7
Other “Sibkabel” load node rated voltage value U BAS 2 6.0 kV was set in “RastrWin” software
package to demonstrate base voltage influence on static load model.
3 Experimental results
Series of steady state mode calculation were made with different load node voltage values, and active
and reactive power values were noted. Obtained results and original polynomial load models are
shown in Fig. 1. Load models difference is explained by using different base voltage values:
U BAS1 6.2 kV for original characteristics and U BAS 2 6.0 kV for “RastrWin” software package.
3
P, MW
Q, Mvar
2.5
Q(U)
U BAS 2
P(U)
2
1.5
1
U BAS1
0.5
U, kV
5
5.2
5.4
5.6
5.8
6
6.2
6.4
6.6
6.8
7
7.2
Figure 1. Original static load models with U BAS1 6.2 kV and “RastrWin” series of steady state mode
calculation results with U BAS 2 6.0 kV
01039-p.2
Smart G rids 2015
Fig. 1 demonstrates impropriety of load node base voltage U BAS changing without base active
PBAS and reactive QBAS power and per unit polynomial coefficients a0 , a1 , a2 , b0 , b1 , b2
corresponding correction.
Correcting expressions were obtained for PBAS , QBAS and coefficients a0 , a1 , a2 , b0 , b1 , b2
calculation on the assumption of absolute unit load models with U BAS1 and U BAS 2 invariance:
2
U BAS (2) U BAS (2)
a2(1) PBAS (2) PBAS (1) a0(1) a1(1) ,
U
U BAS (1)
BAS (1)
a0(2) PBAS (1)
PBAS (2)
a0(1) ,
a1(2) QBAS (2)
b0(2) QBAS (1)
QBAS (2)
(2)
2
PBAS (1) U BAS (2)
a1(1) ,
PBAS (2) U BAS (1)
PBAS (1) U BAS (2) a2(1) ,
U
P
a2(2) BAS (2)
2
U BAS (2) U BAS (2)
QBAS (1) b0(1) b1(1) b2(1) ,
U
U BAS (1)
BAS (1)
b0(1) ,
b1(2)
(3)
BAS (1)
(4)
2
QBAS (1) U BAS (2)
b1(1) ,
QBAS (2) U BAS (1)
b2(2)
QBAS (1) U BAS (2) b2(1) .
U
Q
BAS (2)
(5)
BAS (1)
The equations (2-5) using results are shown in Table 2.
Table 2. Polynomial load model coefficients, base active and reactive power with U BAS 2 6.0 kV
Active power
Reactive power
a0(2)
a1(2)
a2(2)
PBAS(2) , MW
b0(2)
b1(2)
b2(2)
QBAS(2) , Mvar
1.544
-2.297
1.753
1.343
61.614
-124.471
63.857
0.591
Series of steady state mode calculation were repeated with corrected base active PBAS and reactive
QBAS power and per unit polynomial coefficients a0 , a1 , a2 , b0 , b1 , b2 from Table 2. Obtained
results are shown in Fig. 2.
3
P, MW
Q, Mvar
2.5
Q(U)
P(U)
2
1.5
1
0.5
U, kV
5
5.2
5.4
5.6
5.8
6
6.2
6.4
6.6
6.8
7
7.2
Figure 2. Original static load models with U BAS1 6.2 kV and “RastrWin” series of steady state mode
calculation results with U BAS 2 6.0 kV after correction
01039-p.3
MATEC Web of Conferences
In this case, the series of steady state mode calculation results correspond to original static load
models, so presented calculation expressions (2-5) are correct.
4 Conclusion
Polynomial load model practical application in power system studies makes additional demand on
source data input. This demand consists in load node base voltage invariance. Otherwise, static load
models have difference from original characteristics, and calculation complementary error appears. If
base voltage correction is necessary, the base active and reactive power and polynomial coefficients
must be corrected too. Correcting calculation expressions (2-5) are presented in the paper.
References
1.
2.
3.
4.
J. Milanović, K. Yamashita, S. Martínez Villanueva, S. Djokić, and L. Korunović, IEEE
Transactions on Power Systems, 28, 3038 (2013)
W. Price, C. Taylor, G. Rogers, K. Srinivasan, C. Concordia, B. Johnson, IEEE Transactions on
Power Systems, 10, 1302 (1995)
A. Tavlintsev, A. Pazderin, O. Malozemova, P. Chusovitin, Proc. of 13th International
Conference on Environment and Electrical Engineering, (Wroclaw, Poland, 2013)
V.
Neuimin,
“RastrWin”
software
package
user
guide,
http://www.rastrwin.ru/download/Files/HELP_RastrWin3_29_08_12.pdf
01039-p.4
View publication stats
Скачать